1910-11.] Illustration of the Modus Operandi of the Prism. 295 
easily transformed into 
rr\dfJ- 
3X 
by means of the optical equations used 
in § 6. 
§ 8. A general idea can now be obtained from the diagram of the whole 
wave-system outside the prism at the instant the rear of our group has 
reached Q. For clearly, if we repeat what we have done with the group of 
wave-length A, with groups of greater and less wave-lengths we shall find 
that the circle whose centre is O and radius OW is a line of constant 
phase ; and if we draw the isophasal lines through the successive crests of 
our group lying along QW, they must clearly become closer for the shorter 
wave-lengths and open out for the greater, as indicated by the three short 
isophasal lines in the diagram. 
It is evident, also, from the echelon formation of the crests of the wave- 
group which lie along QW in the diagram, why Talbot’s and Powell’s bands 
are obtained only when the retarding plate is inserted from the thin side of 
the prism. The amplitudes of the successive crests of each group within the 
prism could be determined as in the case of ship-waves, though it is pro- 
bable that the number of crests formed by any actual light impulse is very 
small. But, as we have seen in § 7, the grouji in emerging is largely 
extended by the well-known process of the successive wave-crests passing 
through the group, so that we may assume that the amplitudes of the 
successive crests vary slowly as we pass from W to Q in the emergent 
group. 
( Issued separately March 30 , 1911 .) 
